Using Recursive Feature Selection with Random Forest to Improve Protein Structural Class Prediction for Low-Similarity Sequences

Integrated bioinformatics analysis for identifying fibroblast-associated biomarkers and molecular subtypes in human membranous nephropathy

Background

Membranous nephropathy (MN) is characterized by immune complex deposition in the glomerular basement membrane, leading to proteinuria and potentially progressive renal dysfunction. Fibroblasts have been implicated in the pathogenesis of MN through their involvement in tissue remodeling and immune modulation.

Methods

We employed integrated bioinformatics analyses to identify fibroblast-associated biomarkers and molecular subtypes in MN. The xCell algorithm was used to assess fibroblast infiltration, and weighted gene co-expression network analysis (WGCNA) identified fibroblast-related gene modules. Differentially expressed fibroblast-associated genes (DEFAGs) were screened between MN and healthy controls (HC) using differential expression analysis and protein-protein interaction (PPI) network construction. Consensus clustering categorized MN patients into distinct subtypes based on DEFAG expression profiles.

Results

Fibroblast scores were a significant elevation in MN compared to HC, indicating increased fibroblast involvement in MN pathogenesis. WGCNA identified 13 fibroblast-related gene modules, with the brown and turquoise modules showing strong correlation with fibroblast scores (correlation coefficient = 0.79 and 0.75, respectively, p < 0.01). DEFAG analysis revealed 308 genes overlapping between WGCNA and differentially expressed genes (DEGs) in MN. Consensus clustering identified two molecular subtypes (C1 and C2) based on DEFAG expression patterns, with differential gene expression enriching pathways related to immune response and extracellular matrix remodeling. Core biomarker analysis highlighted COL3A1 and TGFB1 as central genes associated with MN, with elevated expression validated across multiple datasets.

Conclusion

Integrated bioinformatics analysis identified fibroblast-associated molecular subtypes in MN, revealing distinct immune profiles and biomarkers. COL3A1 emerged as a potential diagnostic and therapeutic target, implicating its role in immune regulation and disease progression in MN.

Enhancing genomic disorder prediction through Feynman Concordance and Interpolated Nearest Centroid techniques

Clinical biomedical applications of genomic technologies are extensive and provide possibilities to enhance healthcare covering the span of medical talents. Genome disorder prediction is an important issue in biomedical research. Genome disorders cause multivariate diseases such as cancer, dementia, diabetes, Leigh syndrome, etc. Existing machine and deep learning-based methods were introduced to forecast genome disorders. However, the genome prediction outcomes were not sufficient. To address this issue, propose a new method called Quadratic Feynman Polynomial Interpolated and Vector Nearest Centroid-based (QFPI-VNC) for acutely predicting the genome disorder with improved sensitivity and specificity. First, we utilized medical data about children from a public genomes dataset and applied it to Linear Quadratic and Feynman Kac Genome filtering to obtain computationally efficient filtered results. Next, the results are fed to the Concordance Correlated Polynomial Interpolation with the purpose of extracting genome wide data in an accurate manner. Finally, the features extracted are fused and fed to the Support Vector and Nearest Centroid model for genome disorder prediction. Experimental investigations of the proposed method employing the genome dataset confirm that the performance of the proposed method is prospective and in the scope of acceptance with relative to state-of-the-art methods in terms of convergence speed, recognition rate, sensitivity, and specificity. Results suggest that the QFPI-VNC method produces the best performance with a higher genome disease detection rate by 14%, accuracy by 11%, sensitivity by 14% specificity by 12%, and lesser convergence speed by 29% than compared to state-of-the-art methods.

Performance evaluate of different chemometrics formalisms used for prostate cancer diagnosis by NMR-based metabolomics

INTRODUCTION

In general, two characteristics are ever present in NMR-based metabolomics studies: (1) they are assays aiming to classify the samples in different groups, and (2) the number of samples is smaller than the feature (chemical shift) number. It is also common to observe imbalanced datasets due to the sampling method and/or inclusion criteria. These situations can cause overfitting. However, appropriate feature selection and classification methods can be useful to solve this issue.

OBJECTIVES

Investigate the performance of metabolomics models built from the association between feature selectors, the absence of feature selection, and classification algorithms, as well as use the best performance model as an NMR-based metabolomic method for prostate cancer diagnosis.

METHODS

We evaluated the performance of NMR-based metabolomics models for prostate cancer diagnosis using seven feature selectors and five classification formalisms. We also obtained metabolomics models without feature selection. In this study, thirty-eight volunteers with a positive diagnosis of prostate cancer and twenty-three healthy volunteers were enrolled.

RESULTS

Thirty-eight models obtained were evaluated using AUROC, accuracy, sensitivity, specificity, and kappa's index values. The best result was obtained when Genetic Algorithm was used with Linear Discriminant Analysis with 0.92 sensitivity, 0.83 specificity, and 0.88 accuracy.

CONCLUSION

The results show that the pick of a proper feature selection method and classification model, and a resampling method can avoid overfitting in a small metabolomic dataset. Furthermore, this approach would decrease the number of biopsies and optimize patient follow-up. 1H NMR-based metabolomics promises to be a non-invasive tool in prostate cancer diagnosis.

The identification of hub-methylated differentially expressed genes in osteoarthritis patients is based on epigenomic and transcriptomic data

Introduction

Osteoarthritis (OA) refers to a commonly seen degenerative joint disorder and a major global public health burden. According to the existing literature, osteoarthritis is related to epigenetic changes, which are important for diagnosing and treating the disease early. Through early targeted treatment, costly treatments and poor prognosis caused by advanced osteoarthritis can be avoided.

Methods

This study combined gene differential expression analysis and weighted gene co-expression network analysis (WGCNA) of the transcriptome with epigenome microarray data to discover the hub gene of OA. We obtained 2 microarray datasets (GSE114007, GSE73626) in Gene Expression Omnibus (GEO). The R software was utilized for identifying differentially expressed genes (DEGs) and differentially methylated genes (DMGs). By using WGCNA to analyze the relationships between modules and phenotypes, it was discovered that the blue module (MEBlue) has the strongest phenotypic connection with OA (cor = 0.92, p = 4e-16). The hub genes for OA, also known as the hub methylated differentially expressed genes, were identified by matching the MEblue module to differentially methylated differentially expressed genes. Furthermore, this study used Gene set variation analysis (GSVA) to identify specific signal pathways associated with hub genes. qRT-PCR and western blotting assays were used to confirm the expression levels of the hub genes in OA patients and healthy controls.

Results

Three hub genes were discovered: HTRA1, P2RY6, and RCAN1. GSVA analysis showed that high HTRA1 expression was mainly enriched in epithelial-mesenchymal transition and apical junction; high expression of P2RY6 was mainly enriched in the peroxisome, coagulation, and epithelial-mesenchymal transition; and high expression of RCAN1 was mainly enriched in epithelial-mesenchymal-transition, TGF-β-signaling, and glycolysis. The results of the RT-qPCR and WB assay were consistent with the findings.

Discussion

The three genes tested may cause articular cartilage degeneration by inducing chondrocyte hypertrophy, regulating extracellular matrix accumulation, and improving macrophage pro-inflammatory response, resulting in the onset and progression of osteoarthritis. They can provide new ideas for targeted treatment of osteoarthritis.

Predicting the Progress of Tuberculosis by Inflammatory Response-Related Genes Based on Multiple Machine Learning Comprehensive Analysis

Background

Tuberculosis (TB), caused by the bacterium Mycobacterium tuberculosis, affects approximately one-quarter of the global population and is considered one of the most lethal infectious diseases worldwide. The prevention of latent tuberculosis infection (LTBI) from progressing into active tuberculosis (ATB) is crucial for controlling and eradicating TB. Unfortunately, currently available biomarkers have limited effectiveness in identifying subpopulations that are at risk of developing ATB. Hence, it is imperative to develop advanced molecular tools for TB risk stratification.

Methods

The TB datasets were downloaded from the GEO database. Three machine learning models, namely LASSO, RF, and SVM-RFE, were used to identify the key characteristic genes related to inflammation during the progression of LTBI to ATB. The expression and diagnostic accuracy of these characteristic genes were subsequently verified. These genes were then used to develop diagnostic nomograms. In addition, single-cell expression clustering analysis, immune cell expression clustering analysis, GSVA analysis, immune cell correlation, and immune checkpoint correlation of characteristic genes were conducted. Furthermore, the upstream shared miRNA was predicted, and a miRNA-genes network was constructed. Candidate drugs were also analyzed and predicted.

Results

In comparison to LTBI, a total of 96 upregulated and 26 downregulated genes related to the inflammatory response were identified in ATB. These characteristic genes have demonstrated excellent diagnostic performance and significant correlation with many immune cells and immune sites. The results of the miRNA-genes network analysis suggested a potential role of hsa-miR-3163 in the molecular mechanism of LTBI progressing into ATB. Moreover, retinoic acid may offer a potential avenue for the prevention of LTBI progression to ATB and for the treatment of ATB.

Conclusion

Our research has identified key inflammatory response-related genes that are characteristic of LTBI progression to ATB and hsa-miR-3163 as a significant node in the molecular mechanism of this progression. Our analyses have demonstrated the excellent diagnostic performance of these characteristic genes and their significant correlation with many immune cells and immune checkpoints. The CD274 immune checkpoint presents a promising target for the prevention and treatment of ATB. Furthermore, our findings suggest that retinoic acid may have a role in preventing LTBI from progressing to ATB and in treating ATB. This study provides a new perspective for differential diagnosis of LTBI and ATB and may uncover potential inflammatory immune mechanisms, biomarkers, therapeutic targets, and effective drugs in the progression of LTBI into ATB.

Identification of dynamic gene expression profiles during sequential vaccination with ChAdOx1/BNT162b2 using machine learning methods

To date, COVID-19 remains a serious global public health problem. Vaccination against SARS-CoV-2 has been adopted by many countries as an effective coping strategy. The strength of the body’s immune response in the face of viral infection correlates with the number of vaccinations and the duration of vaccination. In this study, we aimed to identify specific genes that may trigger and control the immune response to COVID-19 under different vaccination scenarios. A machine learning-based approach was designed to analyze the blood transcriptomes of 161 individuals who were classified into six groups according to the dose and timing of inoculations, including I-D0, I-D2-4, I-D7 (day 0, days 2–4, and day 7 after the first dose of ChAdOx1, respectively) and II-D0, II-D1-4, II-D7-10 (day 0, days 1–4, and days 7–10 after the second dose of BNT162b2, respectively). Each sample was represented by the expression levels of 26,364 genes. The first dose was ChAdOx1, whereas the second dose was mainly BNT162b2 (Only four individuals received a second dose of ChAdOx1). The groups were deemed as labels and genes were considered as features. Several machine learning algorithms were employed to analyze such classification problem. In detail, five feature ranking algorithms (Lasso, LightGBM, MCFS, mRMR, and PFI) were first applied to evaluate the importance of each gene feature, resulting in five feature lists. Then, the lists were put into incremental feature selection method with four classification algorithms to extract essential genes, classification rules and build optimal classifiers. The essential genes, namely, NRF2, RPRD1B, NEU3, SMC5, and TPX2, have been previously associated with immune response. This study also summarized expression rules that describe different vaccination scenarios to help determine the molecular mechanism of vaccine-induced antiviral immunity.

iEnhancer-DCSV: Predicting enhancers and their strength based on DenseNet and improved convolutional block attention module

Enhancers play a crucial role in controlling gene transcription and expression. Therefore, bioinformatics puts many emphases on predicting enhancers and their strength. It is vital to create quick and accurate calculating techniques because conventional biomedical tests take too long time and are too expensive. This paper proposed a new predictor called iEnhancer-DCSV built on a modified densely connected convolutional network (DenseNet) and an improved convolutional block attention module (CBAM). Coding was performed using one-hot and nucleotide chemical property (NCP). DenseNet was used to extract advanced features from raw coding. The channel attention and spatial attention modules were used to evaluate the significance of the advanced features and then input into a fully connected neural network to yield the prediction probabilities. Finally, ensemble learning was employed on the final categorization findings via voting. According to the experimental results on the test set, the first layer of enhancer recognition achieved an accuracy of 78.95%, and the Matthews correlation coefficient value was 0.5809. The second layer of enhancer strength prediction achieved an accuracy of 80.70%, and the Matthews correlation coefficient value was 0.6609. The iEnhancer-DCSV method can be found at https://github.com/leirufeng/iEnhancer-DCSV. It is easy to obtain the desired results without using the complex mathematical formulas involved.

A model with deep analysis on a large drug network for drug classification

Drugs are an important means to treat various diseases. They are classified into several classes to indicate their properties and effects. Those in the same class always share some important features. The Kyoto Encyclopedia of Genes and Genomes (KEGG) DRUG recently reported a new drug classification system that classifies drugs into 14 classes. Correct identification of the class for any possible drug-like compound is helpful to roughly determine its effects for a particular type of disease. Experiments could be conducted to confirm such latent effects, thus accelerating the procedures for discovering novel drugs. In this study, this classification system was investigated. A classification model was proposed to assign one of the classes in the system to any given drug for the first time. Different from traditional fingerprint features, which indicated essential drug properties alone and were very popular in investigating drug-related problems, drugs were represented by novel features derived from a large drug network via a well-known network embedding algorithm called Node2vec. These features abstracted the drug associations generated from their essential properties, and they could overview each drug with all drugs as background. As class sizes were of great differences, synthetic minority over-sampling technique (SMOTE) was employed to tackle the imbalance problem. A balanced dataset was fed into the support vector machine to build the model. The 10-fold cross-validation results suggested the excellent performance of the model. This model was also superior to models using other drug features, including those generated by another network embedding algorithm and fingerprint features. Furthermore, this model provided more balanced performance across all classes than that without SMOTE.

Mitochondria dysfunction in airway epithelial cells is associated with type 2-low asthma

Background: Type 2 (T2)-low asthma can be severe and corticosteroid-resistant. Airway epithelial cells play a pivotal role in the development of asthma, and mitochondria dysfunction is involved in the pathogenesis of asthma. However, the role of epithelial mitochondria dysfunction in T2-low asthma remains unknown. Methods: Differentially expressed genes (DEGs) were identified using gene expression omnibus (GEO) dataset GSE4302, which is originated from airway epithelial brushings from T2-high (n = 22) and T2-low asthma patients (n = 20). Gene set enrichment analysis (GSEA) was implemented to analyze the potential biological pathway involved between T2-low and T2-high asthma. T2-low asthma related genes were identified using weighted gene co-expression network analysis (WGCNA). The mitochondria-related genes (Mito-RGs) were referred to the Molecular Signatures Database (MSigDB). T2-low asthma related mitochondria (T2-low-Mito) DEGs were obtained by intersecting the DEGs, T2-low asthma related genes, and Mito-RGs. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) was performed to further explore the potential function of the T2-low-Mito DEGs. In addition, the hub genes were further identified by protein-protein interaction (PPI), and the expressions of hub genes were verified in another GEO dataset GSE67472 and bronchial brushings from patients recruited at Tongji Hospital. Results: Six hundred and ninety-two DEGs, including 107 downregulated genes and 585 upregulated genes were identified in airway epithelial brushings from T2-high and T2-low asthma patients included in GSE4302 dataset. GSEA showed that mitochondrial ATP synthesis coupled electron transport is involved in T2-low asthma. Nine hundred and four T2-low asthma related genes were identified using WGCNA. Twenty-two T2-low-Mito DEGs were obtained by intersecting the DEGs, T2-low asthma and Mito-RGs. The GO enrichment analysis of the T2-low-Mito DEGs showed significant enrichment of mitochondrial respiratory chain complex assembly, and respiratory electron transport chain. PPI network was constructed using 22 T2-low-Mito DEGs, and five hub genes, ATP5G1, UQCR10, NDUFA3, TIMM10, and NDUFAB1, were identified. Moreover, the expression of these hub genes was validated in another GEO dataset, and our cohort of asthma patients. Conclusion: This study suggests that mitochondria dysfunction contributes to T2-low asthma.

Comprehensive bioinformatics analysis reveals biomarkers of DNA methylation-related genes in varicose veins

Background: Patients with Varicose veins (VV) show no obvious symptoms in the early stages, and it is a common and frequent clinical condition. DNA methylation plays a key role in VV by regulating gene expression. However, the molecular mechanism underlying methylation regulation in VV remains unclear. Methods: The mRNA and methylation data of VV and normal samples were obtained from the Gene Expression Omnibus (GEO) database. Methylation-Regulated Genes (MRGs) between VV and normal samples were crossed with VV-associated genes (VVGs) obtained by weighted gene co-expression network analysis (WGCNA) to obtain VV-associated MRGs (VV-MRGs). Their ability to predict disease was assessed using receiver operating characteristic (ROC) curves. Biomarkers were then screened using a random forest model (RF), support vector machine model (SVM), and generalized linear model (GLM). Next, gene set enrichment analysis (GSEA) was performed to explore the functions of biomarkers. Furthermore, we also predicted their drug targets, and constructed a competing endogenous RNAs (ceRNA) network and a drug target network. Finally, we verified their mRNA expression using quantitative real-time polymerase chain reaction (qRT-PCR). Results: Total three VV-MRGs, namely Wnt1-inducible signaling pathway protein 2 (WISP2), Cysteine-rich intestinal protein 1 (CRIP1), and Odd-skipped related 1 (OSR1) were identified by VVGs and MRGs overlapping. The area under the curves (AUCs) of the ROC curves for these three VV-MRGs were greater than 0.8. RF was confirmed as the optimal diagnostic model, and WISP2, CRIP1, and OSR1 were regarded as biomarkers. GSEA showed that WISP2, CRIP1, and OSR1 were associated with oxidative phosphorylation, extracellular matrix (ECM), and respiratory system functions. Furthermore, we found that lncRNA MIR17HG can regulate OSR1 by binding to hsa-miR-21-5p and that PAX2 might treat VV by targeting OSR1. Finally, qRT-PCR results showed that the mRNA expression of the three genes was consistent with the results of the datasets. Conclusion: This study identified WISP2, CRIP1, and OSR1 as biomarkers of VV through comprehensive bioinformatics analysis, and preliminary explored the DNA methylation-related molecular mechanism in VV, which might be important for VV diagnosis and exploration of potential molecular mechanisms.

YgiM may act as a trigger in the sepsis caused by Klebsiella pneumoniae through the membrane-associated ceRNA network

Sepsis is one of the diseases that can cause serious mortality. In E. coli, an inner membrane protein YgiM encoded by gene ygiM can target the eukaryotic peroxisome. Peroxisome is a membrane-enclosed organelle associated with the ROS metabolism and was reported to play the key role in immune responses and inflammation during the development of sepsis. Klebsiella pneumoniae (K. pneumoniae) is one of the important pathogens causing sepsis. However, the function of gene vk055_4013 which is highly homologous to ygiM of E. coli has not been demonstrated in K. pneumoniae. In this study, we prepared ΔygiM of K. pneumoniae ATCC43816, and found that the deletion of ygiM did not affect bacterial growth and mouse mortality in the mouse infection model. Interestingly, ΔygiM not only resulted in reduced bacterial resistance to macrophages, but also attenuated pathological manifestations in mouse organs. Furthermore, based on the data of Gene Expression Omnibus, the expression profiles of micro RNAs (miRNAs) and messenger RNAs (mRNAs) in the serum of 44 sepsis patients caused by K. pneumoniae infection were analyzed, and 11 differently expressed miRNAs and 8 DEmRNAs associated with the membrane function were found. Finally, the membrane-associated competing endogenous RNAs (ceRNAs) network was constructed. In this ceRNAs network, DEmiRNAs (hsa-miR-7108-5p, hsa-miR-6780a-5p, hsa-miR-6756-5p, hsa-miR-4433b-3p, hsa-miR-3652, hsa-miR-342-3p, hsa-miR-32-5p) and their potential downstream target DEmRNAs (VNN1, CEACAM8, PGLYRP1) were verified in the cell model infected by wild type and ΔygiM of K. pneumoniae, respectively. Taken together, YgiM may trigger the sepsis caused by K. pneumoniae via membrane-associated ceRNAs. This study provided new insights into the role of YgiM in the process of K. pneumoniae induced sepsis.

A Ferroptosis-Related LncRNA Signature Associated with Prognosis, Tumor Immune Environment, and Genome Instability in Hepatocellular Carcinoma

Background

Ferroptosis is an iron-dependent form of cell death. In this study, we identified ferroptosis-related long noncoding RNAs (FRlncRNAs) to investigate their association with hepatocellular carcinoma (HCC) in prognosis, tumor immune environment, and genome instability.

Methods

Transcriptome profile data of HCC were retrieved from a public database. FRlncRNAs were identified by co-expression analysis. Patients were randomly divided into training and test cohorts. Univariate Cox analysis and Least Absolute Shrinkage and Selection Operator (LASSO) Cox regression were performed to construct a risk model. Patients were divided into high- and low-risk groups based on the risk model. AUC and C index were used to assess the risk model. Survival analysis, immune status, and genome instability were compared between the two groups.

Results

Sixteen FRlncRNAs were identified and used to construct an FRlncRNA signature for the risk model. The Kaplan-Meier analysis revealed that patients in the high-risk group had poorer overall survival than patients in the low-risk group. The area under curve of the risk model was 0.879, 0.809, and 0.757 in the training cohort and 0.635, 0.688, and 0.739 in the test cohort at 1, 2, and 3 years, respectively. The risk model was an independent prognostic predictor and showed excellent prediction of prognosis compared with clinicopathological features. For the differentially expressed ferroptosis-related genes, many enriched metabolic pathways were identified in the functional enrichment analysis. Immune cells such as CD8+ T cells, macrophages M1, natural killer cells, and B cells, which may be associated with antitumor immune responses, differed between the high- and low-risk groups. Genome instability based on the risk model was also explored. A total of 61 genes were differently mutated between the two risk groups, and among them, TP53, HECW2, TRIM66, MCTP2, and KIAA1551 had the most significant mutation frequency differences.

Conclusion

The FRlncRNA signature is closely related with overall survival, tumor immune environment, and genome instability in HCC.

Identification of methylation signatures associated with CAR T cell in B-cell acute lymphoblastic leukemia and non-hodgkin’s lymphoma

CD19-targeted CAR T cell immunotherapy has exceptional efficacy for the treatment of B-cell malignancies. B-cell acute lymphocytic leukemia and non-Hodgkin’s lymphoma are two common B-cell malignancies with high recurrence rate and are refractory to cure. Although CAR T-cell immunotherapy overcomes the limitations of conventional treatments for such malignancies, failure of treatment and tumor recurrence remain common. In this study, we searched for important methylation signatures to differentiate CAR-transduced and untransduced T cells from patients with acute lymphoblastic leukemia and non-Hodgkin’s lymphoma. First, we used three feature ranking methods, namely, Monte Carlo feature selection, light gradient boosting machine, and least absolute shrinkage and selection operator, to rank all methylation features in order of their importance. Then, the incremental feature selection method was adopted to construct efficient classifiers and filter the optimal feature subsets. Some important methylated genes, namely, SERPINB6, ANK1, PDCD5, DAPK2, and DNAJB6, were identified. Furthermore, the classification rules for distinguishing different classes were established, which can precisely describe the role of methylation features in the classification. Overall, we applied advanced machine learning approaches to the high-throughput data, investigating the mechanism of CAR T cells to establish the theoretical foundation for modifying CAR T cells.

A Novel Multistep Iterative Technique for Models in Medical Sciences with Complex Dynamics

This paper proposes a three-step iterative technique for solving nonlinear equations from medical science. We designed the proposed technique by blending the well-known Newton's method with an existing two-step technique. The method needs only five evaluations per iteration: three for the given function and two for its first derivatives. As a result, the novel approach converges faster than many existing techniques. We investigated several models of applied medical science in both scalar and vector versions, including population growth, blood rheology, and neurophysiology. Finally, some complex-valued polynomials are shown as polynomiographs to visualize the convergence zones.

Identifying Functions of Proteins in Mice With Functional Embedding Features

In current biology, exploring the biological functions of proteins is important. Given the large number of proteins in some organisms, exploring their functions one by one through traditional experiments is impossible. Therefore, developing quick and reliable methods for identifying protein functions is necessary. Considerable accumulation of protein knowledge and recent developments on computer science provide an alternative way to complete this task, that is, designing computational methods. Several efforts have been made in this field. Most previous methods have adopted the protein sequence features or directly used the linkage from a protein–protein interaction (PPI) network. In this study, we proposed some novel multi-label classifiers, which adopted new embedding features to represent proteins. These features were derived from functional domains and a PPI network via word embedding and network embedding, respectively. The minimum redundancy maximum relevance method was used to assess the features, generating a feature list. Incremental feature selection, incorporating RAndom k-labELsets to construct multi-label classifiers, used such list to construct two optimum classifiers, corresponding to two key measurements: accuracy and exact match. These two classifiers had good performance, and they were superior to classifiers that used features extracted by traditional methods.

Identification of Type 2 Diabetes Biomarkers From Mixed Single-Cell Sequencing Data With Feature Selection Methods

Diabetes is the most common disease and a major threat to human health. Type 2 diabetes (T2D) makes up about 90% of all cases. With the development of high-throughput sequencing technologies, more and more fundamental pathogenesis of T2D at genetic and transcriptomic levels has been revealed. The recent single-cell sequencing can further reveal the cellular heterogenicity of complex diseases in an unprecedented way. With the expectation on the molecular essence of T2D across multiple cell types, we investigated the expression profiling of more than 1,600 single cells (949 cells from T2D patients and 651 cells from normal controls) and identified the differential expression profiling and characteristics at the transcriptomics level that can distinguish such two groups of cells at the single-cell level. The expression profile was analyzed by several machine learning algorithms, including Monte Carlo feature selection, support vector machine, and repeated incremental pruning to produce error reduction (RIPPER). On one hand, some T2D-associated genes (MTND4P24, MTND2P28, and LOC100128906) were discovered. On the other hand, we revealed novel potential pathogenic mechanisms in a rule manner. They are induced by newly recognized genes and neglected by traditional bulk sequencing techniques. Particularly, the newly identified T2D genes were shown to follow specific quantitative rules with diabetes prediction potentials, and such rules further indicated several potential functional crosstalks involved in T2D.

CNNLSTMac4CPred: A Hybrid Model for N4-Acetylcytidine Prediction

N4-Acetylcytidine (ac4C) is a highly conserved post-transcriptional and an extensively existing RNA modification, playing versatile roles in the cellular processes. Due to the limitation of techniques and knowledge, large-scale identification of ac4C is still a challenging task. RNA sequences are like sentences containing semantics in the natural language. Inspired by the semantics of language, we proposed a hybrid model for ac4C prediction. The model used long short-term memory and convolution neural network to extract the semantic features hidden in the sequences. The semantic and the two traditional features (k-nucleotide frequencies and pseudo tri-tuple nucleotide composition) were combined to represent ac4C or non-ac4C sequences. The eXtreme Gradient Boosting was used as the learning algorithm. Five-fold cross-validation over the training set consisting of 1160 ac4C and 10,855 non-ac4C sequences obtained the area under the receiver operating characteristic curve (AUROC) of 0.9004, and the independent test over 469 ac4C and 4343 non-ac4C sequences reached an AUROC of 0.8825. The model obtained a sensitivity of 0.6474 in the five-fold cross-validation and 0.6290 in the independent test, outperforming two state-of-the-art methods. The performance of semantic features alone was better than those of k-nucleotide frequencies and pseudo tri-tuple nucleotide composition, implying that ac4C sequences are of semantics. The proposed hybrid model was implemented into a user-friendly web-server which is freely available to scientific communities: http://47.113.117.61/ac4c/ . The presented model and tool are beneficial to identify ac4C on large scale.

Drug-Drug Interactions Prediction Using Fingerprint Only

Combination drug therapy is an efficient way to treat complicated diseases. Drug-drug interaction (DDI) is an important research topic in this therapy as patient safety is a problem when two or more drugs are taken at the same time. Traditionally, in vitro experiments and clinical trials are common ways to determine DDIs. However, these methods cannot meet the requirements of large-scale tests. It is an alternative way to develop computational methods for predicting DDIs. Although several previous methods have been proposed, they always need several types of drug information, limiting their applications. In this study, we proposed a simple computational method to predict DDIs. In this method, drugs were represented by their fingerprint features, which are most widely used in investigating drug-related problems. These features were refined by three models, including addition, subtraction, and Hadamard models, to generate the representation of DDIs. The powerful classification algorithm, random forest, was picked up to build the classifier. The results of two types of tenfold cross-validation on the classifier indicated good performance for discovering novel DDIs among known drugs and acceptable performance for identifying DDIs between known drugs and unknown drugs or among unknown drugs. Although the classifier adopted a sample scheme to represent DDIs, it was still superior to other methods, which adopted features generated by some advanced computer algorithms. Furthermore, a user-friendly web-server, named DDIPF (http://106.14.164.77:5004/DDIPF/), was developed to implement the classifier.

Identification of protein functions in mouse with a label space partition method

Protein is very important for almost all living creatures because it participates in most complicated and essential biological processes. Determining the functions of given proteins is one of the most essential problems in protein science. Such determination can be conducted through traditional experiments. However, the experimental methods are always time-consuming and of high costs. In recent years, computational methods give useful aids for identification of protein functions. This study presented a new multi-label classifier for identifying functions of mouse proteins. Due to the number of functional types, which were termed as labels in the classification procedure, a label space partition method was employed to divide labels into some partitions. On each partition, a multi-label classifier was constructed. The classifiers based on all partitions were integrated in the proposed classifier. The cross-validation results proved that the proposed classifier was of good performance. Classifiers with label partition were superior to those without label partition or with random label partition.

Predicting RNA 5-Methylcytosine Sites by Using Essential Sequence Features and Distributions

Methylation is one of the most common and considerable modifications in biological systems mediated by multiple enzymes. Recent studies have shown that methylation has been widely identified in different RNA molecules. RNA methylation modifications have various kinds, such as 5-methylcytosine (m5C). However, for individual methylation sites, their functions still remain to be elucidated. Testing of all methylation sites relies heavily on high-throughput sequencing technology, which is expensive and labor consuming. Thus, computational prediction approaches could serve as a substitute. In this study, multiple machine learning models were used to predict possible RNA m5C sites on the basis of mRNA sequences in human and mouse. Each site was represented by several features derived from $k$ -mers of an RNA subsequence containing such site as center. The powerful max-relevance and min-redundancy (mRMR) feature selection method was employed to analyse these features. The outcome feature list was fed into incremental feature selection method, incorporating four classification algorithms, to build efficient models. Furthermore, the sites related to features used in the models were also investigated.

Identification of Pan-Cancer Biomarkers Based on the Gene Expression Profiles of Cancer Cell Lines

There are many types of cancers. Although they share some hallmarks, such as proliferation and metastasis, they are still very different from many perspectives. They grow on different organ or tissues. Does each cancer have a unique gene expression pattern that makes it different from other cancer types? After the Cancer Genome Atlas (TCGA) project, there are more and more pan-cancer studies. Researchers want to get robust gene expression signature from pan-cancer patients. But there is large variance in cancer patients due to heterogeneity. To get robust results, the sample size will be too large to recruit. In this study, we tried another approach to get robust pan-cancer biomarkers by using the cell line data to reduce the variance. We applied several advanced computational methods to analyze the Cancer Cell Line Encyclopedia (CCLE) gene expression profiles which included 988 cell lines from 20 cancer types. Two feature selection methods, including Boruta, and max-relevance and min-redundancy methods, were applied to the cell line gene expression data one by one, generating a feature list. Such list was fed into incremental feature selection method, incorporating one classification algorithm, to extract biomarkers, construct optimal classifiers and decision rules. The optimal classifiers provided good performance, which can be useful tools to identify cell lines from different cancer types, whereas the biomarkers (e.g. NCKAP1, TNFRSF12A, LAMB2, FKBP9, PFN2, TOM1L1) and rules identified in this work may provide a meaningful and precise reference for differentiating multiple types of cancer and contribute to the personalized treatment of tumors.

A Deep Learning and XGBoost-Based Method for Predicting Protein-Protein Interaction Sites

Knowledge about protein-protein interactions is beneficial in understanding cellular mechanisms. Protein-protein interactions are usually determined according to their protein-protein interaction sites. Due to the limitations of current techniques, it is still a challenging task to detect protein-protein interaction sites. In this article, we presented a method based on deep learning and XGBoost (called DeepPPISP-XGB) for predicting protein-protein interaction sites. The deep learning model served as a feature extractor to remove redundant information from protein sequences. The Extreme Gradient Boosting algorithm was used to construct a classifier for predicting protein-protein interaction sites. The DeepPPISP-XGB achieved the following results: area under the receiver operating characteristic curve of 0.681, a recall of 0.624, and area under the precision-recall curve of 0.339, being competitive with the state-of-the-art methods. We also validated the positive role of global features in predicting protein-protein interaction sites.